Looking for criteria to assess the relevance of structural flexibility on the response of large buried structures subject to seismic action

Abstract

This work analyzes the requirements of the models needed to estimate the seismic motions observed along large cylindrical buried structures by performing a parametric analysis of the problem using two models: one in which the structure is considered as perfectly rigid, and another one in which its actual structural flexibility is taken into account. The properties of the soil, the flexibility of the structure and the variability of the seismic incident field along the buried length are the three key aspects that affect the seismic response of the system. The parametric analysis has been carried out using a wide range of properties for both, structure and soil. Thus, computing the seismic response of a relatively large number of configurations is needed and it makes advisable the use of a numerical tool of low computational cost but accurate enough. This is why the study is performed using two models based on a Beam-onDynamic-Winkler-Foundation approach. These two models were previously verified by comparison against results obtained for the problem at hand using a more rigorous 3D multidomain boundary element model. The amount of results obtained by comparison of the seismic responses estimated by both models is significantly large and needs to be synthesized. These results are used to build and propose a specific criterion that can be used to elucidate under which circumstances is it possible to neglect the structural flexibility. It is found that, contrary to what is commonly assumed, the structural slenderness ratio alone cannot be used, in general, to predict the validity of the rigid structure approach: embedment lengths, soil stiffness, depth of interest and natural period of study are, also, key parameters that need to be taken into account. A close-form criterion is proposed in table form taking all such parameters into account.